Is Arm Dominance Associated With Clinically Meaningful Differences in Outcomes After Shoulder Arthroplasty?
Academic Article
Overview
abstract
BACKGROUND: Limited available evidence seems to suggest that the increased use of the dominant (versus nondominant) limb may allow for earlier return to function and better ROM in the dominant limb at 12-month follow-up after anatomic or reverse total shoulder arthroplasty (TSA). Nevertheless, whether the earlier achievement of physical therapy milestones is associated with a clinically meaningful difference in patient-reported outcome measures (PROMs) is yet to be determined. QUESTIONS/PURPOSES: (1) What are the 12-month minimum clinically important difference (MCID) and patient acceptable symptom state (PASS) thresholds for the Shoulder Pain and Disability Index (SPADI), QuickDASH, numeric rating scale (NRS) for pain, and Constant-Murley score? (2) Is there a difference between the dominant- and nondominant-side TSAs in terms of the proportions of patients achieving an MCID or PASS at 12-month follow-up? METHODS: This retrospective, comparative study analyzed data from a longitudinally maintained shoulder arthroplasty registry at a specialized orthopaedic institution. Patients were eligible for inclusion if they underwent primary anatomic or reverse TSA from 2006 to 2024 for cuff tear arthropathy or primary osteoarthritis and had 12-month follow-up for at least one PROM. We collected relevant baseline patient-related and procedure-related characteristics. The main association of interest was operated limb relative to limb dominance, and shoulders were stratified into the dominant-side or nondominant-side group. A total of 2152 shoulders, 65% (1404) of which were in the dominant-side group, were analyzed. The mean age was 73 years, and the majority of patients were women in both groups. In the dominant group, more patients were treated for cuff tear arthropathy, and a larger proportion received a reverse TSA. Loss to follow-up at 12 months did not differ between groups, reaching 13% for the dominant group and 16% for the nondominant group. The PROMs collected were the SPADI, QuickDASH, NRS for pain, and Constant-Murley score. These were administered at baseline and at 12 months postoperatively. The MCID and PASS thresholds for the PROMs of interest were estimated using a distribution-based approach. A sensitivity analysis was performed using the best available evidence for anchor-based MCIDs (20 for the SPADI, 12 for the QuickDASH, 2.2 for the NRS for pain, and 9 for the Constant-Murley score). Adjusted comparisons of distribution-based MCID estimates and PASS proportions between dominant- and nondominant-side procedures were conducted using generalized linear mixed-effects logistic regression models. Models were adjusted for admission type, surgical indication, procedure type, cuff tear severity, BMI, baseline ROM (forward flexion, abduction, external rotation), and baseline QuickDASH, with a random intercept for patient ID to account for within-patient clustering. Results are reported as ORs with 95% confidence intervals (CIs). RESULTS: Distribution-based absolute MCID estimates were 16 for SPADI, 13 for QuickDASH, 1.5 (reduction) for NRS for pain, and 12 for Constant-Murley score. Distribution-based absolute PASS estimates were 19 for SPADI, 5 for QuickDASH, 1.8 (reduction) for NRS for pain, and 20 for Constant-Murley score. At 12 months, MCID and PASS responder proportions did not differ in clinically important ways between dominant and nondominant shoulders, but approximately 25% (509 of 2120) of shoulders did not achieve the MCID for the QuickDASH, more than 10% (247 of 2123) did not achieve it for pain, and approximately 15% did not achieve a PASS for one or more outcomes tool. All absolute risk differences and 95% CIs fell within the prespecified ± 10% smallest important difference margin, indicating no clinically important dominance effect. In adjusted mixed-effects logistic regression models, dominance was not associated with meaningful differences in the odds of achieving MCID or PASS for SPADI, QuickDASH, NRS for pain, or Constant-Murley score (ORs near 1.0, with all CIs crossing unity). Across all sensitivity analyses, arm dominance demonstrated no association with 12-month MCID or PASS after TSA. CONCLUSION: Arm dominance was not meaningfully associated with an increased or decreased likelihood of achieving MCID or PASS across multiple PROMs. However, as many as 1 in 4 patients did not reach MCID or PASS thresholds after TSA, irrespective of arm dominance. These findings indicate that, in typical clinical practice, dominance should not be considered a major determinant of postoperative recovery expectations. LEVEL OF EVIDENCE: Level III, therapeutic study.
